KR0170354B1 - Television signal processing system for reducing diagonal image artifacts - Google Patents

Abstract

The present invention relates to an apparatus for reducing or eliminating unnecessary oblique processed images in any image displayed by a system using an on-screen signal processing technique.

Description

[Name of invention]

Television signal processing device for reducing the oblique process image

[Background of invention]

The present invention relates to an apparatus for reducing or eliminating unnecessary diagonal processed images, for example, in any image displayed by a system using an on-screen signal processing technique.

Conventional TV receivers, such as receivers in accordance with the NTSC broadcast standard, adopted in the United States and elsewhere, have an aspect ratio of 4: 3 (ratio of width to height of the displayed image).

Recently, attention has been drawn to using relatively high aspect ratios for TV receiver systems such as 2: 1, 16: 9 or 5: 3. Video information signals with a 5: 3 aspect ratio require special attention when this ratio approaches the ratio of moving picture film. Therefore, such a signal is transmitted or received without cutting the image information. Optical screen TV systems that transmit signals with increased aspect ratios as compared to conventional systems are not compatible with receivers with conventional aspect ratios. This makes it difficult to adopt a system with a wide screen.

Therefore, it is desirable to have a wide screen system that is compatible with existing TV receivers. Such a system is disclosed on November 1, 1988, in M.A. The name of the invention granted by Isnardi is disclosed in patent 4,782,383, a device for processing high frequency edge information in optical screen TV systems.

It is above all desirable to have a wide screen system with facilities for improving the sharpness of the displayed image. This is to provide special image clarity. For example, such a wide screen EDTV system may have an apparatus for providing a highly scanned image. Such a system is published in a newspaper entitled Encoding for Retrievability and Suitability in ACTV Systems. This paper, IEEE Report on Broadcasting Vol. BC-33, No. 4, issued to Isnardi and other doctors, published on pages 106-123 in December 1987.

It has been observed that intra-picture video image information processing may result in a reduction of the diagonal edge image and produce unnecessary distorted processed images. In accordance with the principles of the present invention, an apparatus is described which reduces the oblique processed image here.

[Summary of invention]

The apparatus according to the principles of the present invention comprises, by way of example, means for suitably processing video signals that are susceptible to display unwanted crushed diagonally processed images. The distorted diagonal processed image is caused by in-screen signal processing.

The information sample of the moving image in any encoder, i.e. in the described embodiment of the invention, receives for example 1.5 MHz field repeats for a given frequency range. At the decoder, the information sample of the stationary image is subjected to a frame repeat process in a given field repeat frequency range. On the other hand, an information sample of moving information remains unchanged.

The selective processing of moving or floating image information at the encoder and decoder is controlled for any auxiliary signal element, where the auxiliary signal element contains field difference information and is a sample of existing or removed image movement.

The apparatus is described in the text of a wide screen EDTV television system using an intra picture signal processing method. The widescreen EDTV signal contains a dual element having a central panel and a first major element comprising time-compressed side panel information and a second auxiliary element containing side panel information.

In the main device, only the center panel information is subjected to intra frame processing.

In the disclosed compatible wide screen EDTV television system, the original high definition, ie very actively scanned wide screen signal was encoded to include four elements. The above four devices are each processed before they are recombined within a single signal transmission channel.

The first element is the main signal scanned alternately at 2: 1 with an aspect ratio of standard 4: 3. The device constitutes the central portion of the wide picture signal, i.e. its portion extended to occupy almost 4: 3 aspect ratio active line time, and also constitutes a side panel of horizontal low-frequency information, which is horizontal and overscanning of the horizontal image. The area, ie, the information, is compressed in an area hidden within the standard television receiver display. Only the center part of the device is susceptible to intra frames that average over a given frequency.

The second element is an auxiliary signal that scans alternately 2: 1 constituting the left and right panel high frequency information, and the information is a time extended to 1/2 of the active line time.

The extended side panel information thus occupies mainly the entire total line time. This element is arrange | positioned so that it may occupy the same time as the center part of a primary element. And an averaged intra frame.

The third element is obtained from a wide screen signal source as an auxiliary signal alternately scanned at 2: 1 and has high frequency horizontal emission information between about 5.0 MHz and 6.0 MHz. The device is also arranged to occupy the same period as the central part of the first device. And an averaged intra frame. The averaged intra frame second and third element quadratures adjust the phase to control the alternating subcarriers. In this case, the alternating subcarriers are combined with an intra frame that averages the first elements. The fourth element is a 2: 1 secondary interlaced scanned tidal signal that constitutes the luminance detail of the temporary potential difference, which helps to reconstruct the missing image information in the widescreen EDTV receiver.

The fourth device includes image motion information. At this time, the appropriate signal processing apparatus announced responds to this information in order to reduce unnecessary diagonal processed images.

In a wide-screen EDTV receiver, a composite signal containing four devices described previously is decoded as a member of the above devices. These decoded elements are each processed and used to develop a video display widescreen signal with improved clarity.

[Brief Description of Drawings]

1 is a general explanatory view of a combined optical screen EDTV encoder system comprising a device according to the invention.

La is a detailed block diagram of an encoder for a system of the present system.

1B-1E are diagrams to help understand the principle of operation of the present system.

2-5 show signal waveforms and diagrams to help understand the principle of operation of the system.

13 is a block diagram of a portion of an optical screen EDTV receiver decoder device.

6 to 12 and 14 to 27 illustrate the aspect ratio of the present system in more detail.

Detailed description of the invention

The system, which is intended to transmit a wide aspect ratio image, for example a ratio of 5: 3, over this standard NTSC broadcast channel, must obtain a good picture display by a wide receiver. The system, on the other hand, eliminates or reduces the noticeable degradation in a 4: 3 standard aspect ratio display. The use of signal compression techniques on the side of the picture uses a standard NTSC TV receiver display. However, this may impair the image sharpness in the side region of the reconstructed wide screen image.

Since temporal compression results in an extension of the frequency domain, only low frequency devices will withstand processing in standard TV channels.

And this shows a relatively narrow bandwidth when compared to the bandwidth required for a wide picture signal. Therefore, when the compressed side of a suitable wide screen signal is compressed onto a wide screen receiver, a significant difference is caused between the high frequency component or the sharpness of the side panel and the center of the displayed widescreen image, and otherwise, this step is to be avoided. Steps are carried out.

This significant difference is caused by the recovery of low frequency side panel information. However, high frequency information is lost by the effect band limited video channel.

In the system of FIG. 1, the elements common to the more precise system of FIG. La are the same with the same reference numerals.

As shown in FIG. 1, the original optical screen progressive scan signal with left, right and center panel information is processed to advance four encoding components. These four elements have been described above and illustrated schematically in FIG. The processing of the first element (including the extended time center and the side portion of the time compressed low frequency information) is that the resulting luminance band does not exceed the NTSC luminance band of 4.2 MHz in this example. The signal is a color encoded in the standard NTSC format, and the luminance and chrominance components of the signal are moderately prefiltered (using a field comb filter), which is an improved luminance chrominance separation for standard NTSC and optical screen receivers. To provide.

The time extension of the second device (side panel of high frequency information) reduces its horizontal bandwidth to about 1.16 MHz. This device is not spatially correlated with the main signal. Special care must be taken to mask the sharpness of the device on a standard NTSC receiver.

The high frequency luminance information component of the third component extending from 5.0 to 6.0 MHz is first shifted to the lower frequency for the frequency range 0 to 1.0 MHz before processing.

The fourth device (momentary field difference helper) is arranged in 4: 3 standard format to associate the main signal component with the present device to mask the visibility of the device on a standard NTSC receiver. And the fourth device horizontally limits the bandwidth to 750 KHz.

In more detail, the first, second, and third devices are processed by respective intra frame averagers 38, 64 and 76 (VT filter type), which are combined with the main signal in the optical screen receiver. This is to eliminate VT crosstalk between auxiliary signal elements. The center panel information of the first signal is an intra frame averaged over about 1.5 MHz.

The second and third intra frame average devices, denoted X and Z, are modulated 3.108 MHz alternating subcarriers with compressed atypical amplification prior to quadrature and have a field alternating phase within block 80. The modulated signal from block 80 is applied to intra frame average first element in adder 40.

The resulting output signal is a 4.2 MHz band baseband signal (NTSCF). This modulates the RF picture carrier in block 57 with the low pass filter 750 KHz of the fourth element from filter 79 to provide an NTSC RF signal, which is a widescreen scan receiver over a single standard band broadcast channel. Or can be sent to a standard NTSC receiver.

The use of temporal compression on the first element completely compresses low frequency side panel information into the horizontal and scanning regions of a standard NTSC signal. The high frequency side panel information of the second element and the high frequency luminance detail information of the third element coexist spectrally with the standard NTSC signal through the video transmission channel, which method alternates subcarrier quadrature including block 80 to be discussed later. It is according to the modulation technique.

When received by a standard NTSC receiver, only the center panel portion of the main signal is visible. The second and third devices can make a low amplification interference fringe. This interference fringe is not noticeable at normal viewing distance and normal image control settings. The fourth element is completely displaced within the receiver with a tuned video detector. In a receiver with an inclusion detector, the fourth element is processed but not recognized since this element relates to the main signal.

The main signal shows approximately 52 standard NTSC active horizontal line intervals. Only high frequency information of 1.5 MHz or higher of the present device is an averaged intra frame.

The time-compressed side panel low frequency information of the device is difficult to undergo intra frame averaging. This selective intra-frame processing of the primary device can improve the clarity of the diagonal side panel image information by eliminating unnecessary distorted diagonally processed images, sometimes called jazzy, which is called if the compressed side panel information of the primary signal If it is an averaged intra frame, it may be provided in the reconstructed image.

In this respect, the low frequency information of the side panel of the main signal element has been time compressed to about 6 due to the side compression element (SCF). If this temporal condensation information is an intra frame averaged before there is a time extension at the receiver for image reconstruction, the reconstructed side panel image information will represent a distorted diagonal. This is because the start of intra frame averaging at the horizontal frequency results in a lower SCF time than for the center panel. Diagonal image information is severely distorted as the frequency at which intra frame averaging is performed decreases.

For example, if the main signal is an intra frame averaged over a frequency above 1.5 MHz, and the first element side panel low frequency information is compressed at an SCF of 6, then the intra frame averaging of the side panel information is at a very low frequency of 250 KHz. Starts, and results in a distorted diagonal.

The warped diagonals therefore become more pronounced within the reconstructed side panel area.

Since the device 1 is not an intra frame averaged in a time-compressed side panel region, the entire region of the original frequency in this region (0 to 700 KHz) retains full vertical sharpness by the distorted diagonal workpiece.

However, the intra frame averages made on elements 2 and 3 as well as the intra frame averages realized on element 1 in the center panel region may result in unnecessary distorted diagonal workpieces, which are reduced as described below. .

The element 2 including the left and right side panel high frequency information is arranged so that the present element occupies the same time interval as the center panel portion of the element 1. Therefore, the left and right side panel high is the time extended to fill the complete center panel area, so that device 2 shows an active horizontal scanning interval of about 50, which is the horizontal of the center panel portion of device 1. Coincides with the scan interval. For this purpose, the lateral extension element SEF is about 4.32 when compared to the SEF of about 4.49 required to extend the left and right side panel information of the device 2 to 52 full active line times.

The elements 2 and 3 are arranged in the central panel area due to the intra frame process carried out on the main element 1 and the auxiliary elements 2 and 3. As described later, intra frame averaging is a process of performing separation such as the auxiliary modulated signal M and the main signal N in this example. Since the intra frame processing region in element 1 of the two combined signal elements is preferentially reduced to include only 50 center panel regions, the mapping of modulation elements 2 and 3 is modulated to include only those center panel regions. .

As mentioned above, the element 3 is arranged to occur simultaneously with the center panel spacing by the time of compressing the extended horizontal luminance information to 50. The element 3, which is compressed from 52 to 50, sacrifices some spatial relationship with the main element 1. But more importantly, the device 3 ensures that the center and side panel areas of the reconstructed image exhibit similar horizontal clarity.

Although the spatial relationship between elements 1 and 3 will mask the effect of crosstalk between the alternating subcarriers and the main signal, the importance of maintaining the complete spatial relationship of the element 3 is that the alternating subcarriers have already formed within the element 2. Decrease because it contains non-associative information. The amount of spatial associations abandoned in the device 3 is negligibly small and consequently becomes important by similar center and side panel horizontal sharpness. The element 4 remains unchanged while showing 52 active line times corresponding to the main signal, while not being an averaged intra frame.

In the decoder, intra frame processing is performed only for the center panel region for separating signals N and M as described in FIG. 13.

After device M has been demodulated into devices 2 and 3, devices 2 and 3 are placed in their original time slit to occupy 52 full active line spacing.

Figure 1b shows the RF spectrum of a published EDTV widescreen system and includes auxiliary information when compared to the RF spectrum of a reference NTSC system.

In the spectrum of the published system, the side panel high and extra high frequency horizontal luminance detail information extends approximately 1.16 MHz on both sides of the 3.1108 MHz alternating subcarrier (ASC).

The V-T assistance signal information (element 4) extends to 750 KHz on both sides of the main signal image carrier frequency.

The optical screen progressive scan receiver includes an apparatus for reconstructing the original optical screen progressive scan signal. Compared to the standard NTSC signal, the reconstructed optical screen signal has a left and right side panel with standard NTSC clarity and a 4: 3 aspect ratio center panel with excellent horizontal and vertical luminance details, especially in still image portions. .

Two basic concepts govern the development of signal processing techniques and the processing of the first, second, third and fourth signal elements. This concept is compatible with existing receivers and resilient to that receiver. Full compatibility refers to the compatibility of receivers and transmitters where existing standard receivers can receive optical screen EDTV signals and provide standard displays without special adapters. For example, in this sense compatibility requires that the transmitter image scanning format be the same as the receiver image scanning format within the same error tolerance. Compatibility also means that non-standard devices are physically hidden within the main signal when displayed on a standard receiver. In order to achieve compatibility by the latter meaning, the published system uses the following technique to conceal the auxiliary device.

As mentioned above, the side panel rows are physically hidden within the normal horizontal area of the standard receiver. When compared to element 2, that is, the side panel low element, this low energy and element 3, i.e., a normally low energy high frequency detail signal, have interlaced frequency (an odd multiple of 1/2 horizontal line rate). Spherically modulated with alternating subcarriers at 3.108 MHz and amplitude compressed. The frequency phase and amplitude of an alternating subcarrier is determined by the visibility of the modulated alternating subcarrier signal. For example, the phase of an alternating subcarrier from field to field differs from the phase shift of the chromatic subcarrier from one field to the next. The sensitivity is selected as much as possible by controlling the phase of the alternating subcarriers from one field to the next to be shifted 180 degrees into the field. Although the modulated alternating subcarrier component is completely within the chromaticity pass band (2.0 to 4.2 MHz), the modulated alternating subcarrier component is hidden because it is represented as a field rate complementary color flicker. This cannot be detected by the human eye at normal levels of chromatic saturation. In addition, nonlinear amplitude suppression of the modulating components prior to modulating the amplitude serves to lower the instantaneous amplitude overshoot to an acceptable low level.

Component 3 is spatially related to the central information portion of component 1 and slightly less spatially related to that associated with the left and right information portions of component 1. This is accomplished by the format encoder as described below.

Component 4, i.e., the helper signal, is also hidden by temporally extending the center panel information to conform to the standard 4: 3 format, so that component 4 is correlated with the main signal.

Component 4 is removed from a standard receiver equipped with a sync detector and since it is spatially correlated with the main signal, it is hidden from a standard receiver equipped with an envelope detector and cannot be detected by the eye.

Recovery of components 1, 2 and 3 in the progressive scanning receiver of the optical screen can be accomplished using intra frame processing at the transmitter and receiver. This process involves the elements 38, 64 and 76 in the transmitter system of FIGS. 1 and 1A and the associated elements in the receiver described below. Intra-frame equalization has two visible correlations for mutual coupling so that signals can be recovered effectively and accurately by field storage without VT (vertical-temporal) crosstalk, even when there is motion in the case of video display signals. It is a signal conditioning technique to prepare a signal. The type of signal conditioning used for this purpose essentially involves creating two identical signals for the field reference by obtaining two samples having the same value. Intra frame leveling is a convenient technique to accomplish this goal, but other techniques may be used. Intra frame leveling is basically a linear time varying digital pre-filtering and post-filtering process to consolidate the correct recovery of the combined signal with two visible correlations. Horizontal crosstalk can be eliminated by guardbanding between the horizontal pre-filter at the transmitter encoder and the post-filters at the receiver decoder.

Intra frame leveling is a form of pair (group) pixel processing. The processing of intra frame leveling in the time domain is shown schematically in FIG. 1C. In this figure the field pairs are made identical by leveling the pixels A, B and C, D 262H apart. The mean value replaces the original value in each pair of groups. 1d illustrates the intra frame leveling process in the system of FIG. Pairs of picture elements 262H apart in one frame starting at components 2 and 3 are averaged and the mean value (e.g., X1, X3 and Z1, Z3) replaces the original pixel value. This V-T leveling occurs within one frame but does not cross the frame boundary. In the case of component 1, intra frame averaging is performed only for center panel information of approximately 1.5 MHz or more in order not to affect low frequency vertical detail information. In the case of components 1 and 2, intra frame equalization is performed on the composite signal including the chromaticity (C) component and the luminance (Y) component for the entire chromaticity band. The chromaticity component of the composite signal survives intra frame equalization because pixels 262H apart are in phase with respect to the color subcarrier. The phase of the new alternating subcarriers is exactly different from the phase for the pixels 262H apart and is controlled not to be the same as the phase of the color subcarriers.

So when components 2 and 3 (after spherical modulation) are added to component 1 in unit 40, the pixels 262H apart have the form (M + A) and (MA), where M is the main synthesized signal above 1.5 MHz. And A is a sample of the auxiliary modulated signal.

Intra-frame equalization eliminates V-T crosstalk even when there is motion.

The process of intra frame leveling produces the same samples 262H apart. It is a simple matter to recover the information content of non-crosstalk samples by processing pixel samples 262H apart in one frame so that the primary and auxiliary signal information is recovered at the receiver.

The intra frame average unique information at the decoder in the receiver can actually be completely recovered through intra frame processing because the unique visible correlation information is actually made of the same field-to-field.

Also at the receiver, the RF channel is spherically modulated using a synchronous RF detector. So component 4 is separated from the other three components. Intra frame processing is used to separate component 1 from modulated components 2 and 3, and spherical modulation is used to separate components 2 and 3 as will be described with reference to FIG.

After the four components are recovered, the composite signal is NTSC decoded and separated into luminance and chroma components. Inverse mapping is performed for all components to recover a wide screen aspect ratio, and the side panels are combined such that the full side panel resolution is restored. The extended high frequency luminance detail is shifted to its original frequency range and added to the luminance signal, which is converted into a progressive scan format using temporal interpolation and helper signals. Chromaticity signals can be converted to progressive scan formats using temporary, unassisted interpolation. Eventually the luminance and chromatic progressive scan signals are converted to analog form and matrixed so that an RGB color image signal is generated for display by a wide screen progressive scan display.

Intra frame leveling processing performed at the encoder can result in undesirable jagged diagonal image artifacts in the form of jagged strata, especially in the distinct light-dark transition regions. The visibility of these artifacts can be significantly reduced by suitably modifying the intra frame-averaging process performed by units 38, 64, and 76 in the encoder, depending on whether or not there is image motion. The processing of the center and side panel information for the purpose of reducing jagged diagonal artifacts is illustrated in FIGS. 25 and 26.

FIG. 25 shows a part of the video signal processed in the presence of motion of the image information in the encoder in its original form, and shows the video signal processed in the presence of the still image information. FIG. 26 shows how a video signal is processed in a decoder in the presence of image information.

The video signal is illustrated for radix and even frames having radix field 1 and even field 2, respectively, in their original and processed form. The odd field in the odd and even frames includes the image pixels A1, C1 and A2, C2, respectively. The even fields in the odd and even frames include pixels B1, D1, and B2, D2, respectively.

When there is motion of the image, the networks 38, 64, and 76 of FIG. 1 average the video signal intra frame over a frequency range of approximately 1.5 MHz to 4.2 MHz, as shown in the central diagram of FIG. The processing of intra frame averaging has been discussed in detail above. In the presence of still picture information, networks 38,67 and 76 perform field repetition substitution over a frequency range of approximately 1.5 MHz to 3.1 MHz. In particular, in this example, the radix field pixel samples for the radix frame are delivered in pairs 262H apart with the radix field pixel values (eg A1 and C1) replacing the associated even field pixel values (eg B1 and D1). . The original pixel values are designated as black dots and the replacement pixel values are designated as white dots.

Similarly, for even frames, even field pixel samples are delivered in pairs 262H apart with even field field values B2 and D2 replacing the associated odd field pixel values (which are A2 and C2, respectively).

In the decoder, frame repetition replacement is performed in the case of a still picture as illustrated in FIG. However, the intra frame averaged signal from the encoder is not disturbed at the decoder in the case of motion pictures. For still images, frame repetition replacement is performed over the same frequency range (1.5 MHz to 3.1 MHz) as field repetition replacement is performed at the encoder. The entire vertical detail is recovered in the still picture area by repeating each of the original pixel values, for example A1, C1, B2, D2, proceeding into the same spaced position within the next frame 525H apart as indicated by the solid arrows.

Alternatively, the original even field pixel values (e.g., B2, D2) may be repeated into equally spaced positions in the preceding field as indicated by the dashed arrows. In this case, frame repeat replacement will be completed within four consecutive fields, which would be convenient for example for tape editing or film editing. The information content of the component 4 helper signal can be conveniently used to determine the presence or absence of image motion in order to control the above-described fit processing operation.

When the chromaticity information band is limited to 500KHz and intra frame equalization, field repetition, and frame repetition are executed over the frequency range, jagged diagonal artifacts are reduced or eliminated to 3.1 MHz, some chroma signal processing artifacts are observed, and chromaticity band Field leveling at s results in improved luminance-chromatic separation at the decoder.

Jagged diagonal image artifacts can be reduced with unsuitable signal processing, but this processing can produce its own artifacts. In a nonconforming scheme, blocks 38, 64, and 76 of FIG. 1 may be replaced with blocks that execute only the field repeat replacement function shown by the diagram on the right of FIG. For even frames, even field samples are sent in pairs, and for odd frames, even field samples are sent in pairs. By using the frame repetition replacement process shown in FIG. 26 at the decoder, leveling may not be performed in the still area and the entire vertical detail may not be recovered.

Non-conforming approaches may generate motion artifacts in standard NTSC receivers. However, the appearance of these motion artifacts can be reduced by V-T filtering of the original optical screen progressive scan signal, such as by the V-T filter network 16 in the system of FIG. If desired, an auxiliary linear combination of pixel values can be used to provide a good exchange between motion artifacts in a standard NTSC receiver and jagged diagonal artifacts in an optical screen receiver. For example, 25% of the even field pixel value and 75% of the even field pixel value 262H apart in the field repeat replacement process may be transmitted for even frames, and 25% and the odd of the even field pixel value 262H apart in the odd frame. 75% of the field pixel values can be transmitted.

The non-fit method can generate motion artifacts in the moving part of the optical screen image while restoring the entire vertical detail in the still part of the optical screen image.

The motion adaptation process in the discussed encoder improves the motion rendition in standard NTSC and optical screen receivers.

Before discussing the combined optical screen encoding system of FIG. 1A, the signal waveforms A and B of FIG. 2 will be discussed. Signal A is a 5: 3 aspect ratio optical screen signal that is converted into a standard NTSC compatible signal with an aspect ratio of 4: 3 as depicted by signal B. The optical screen signal A includes a center panel portion associated with the first image information occupying the interval TC and a left and right panel portion associated with the second image information occupying the interval TS. In this example, the left and right panels exhibit the same aspect ratio that is less than the aspect ratio of the dominant center panel between them.

Optical screen signal A is converted to NTSC signal B by fully compressing some panel information into a horizontal overscan area associated with time interval TO. The standard NTSC signal has an overscan interval TO, a display time interval TD including the video information being displayed, and an active line interval TA (52.6 microsecond duration) surrounding the entire horizontal line time interval TH for 63.556 microseconds. . Intervals TA and TH are the same for optical screens and standard NTSC signals. Almost all consumer television receivers have an overscan interval that accounts for at least 4% of the total active line time TA, ie 2% on the left and right. At an interlaced sampling rate of 4xfsc (where fsc is the frequency of the color carrier), each horizontal line interval includes 910 pixels, of which 754 pixels constitute active horizontal line image information that is displayed.

The optical screen EDTV system is shown in more detail in FIG. Referring to Figures 1A, a 525 line, 60 field / sec wide screen progressive scan camera 10 provides in this example an R, G, B component and a wide aspect ratio of 5: 3 for the optical screen color signal.

Interlaced scan signal sources can be used, but progressive scan signal sources yield much better results. Wide screen cameras have a larger aspect ratio and video bandwidth compared to standard NTSC cameras, and the video bandwidth of an optical screen camera is proportional to the product of its aspect ratio and the total number of lines per frame, among other factors. Assuming constant speed scanning by an optical screen camera, the increase in aspect ratio results in a corresponding increase in video bandwidth as well as horizontal suppression of the image information when the signal is displayed by a standard television receiver having an aspect ratio of 4: 3. Cause. For these reasons, optical screen signals need to be transformed for full NTSC compatibility.

The color video signal processed by the encoder system of FIG. 1 includes luminance and chroma signal components. The luminance and chroma signals include low and high frequency information, so the following discussion will refer to these information as lows and highs, respectively.

The wide bandwidth optical screen progressive scan color video signal from the camera 10 is matrixed in the unit to derive the luminance component Y and the chrominance signal components I and Q from the R, G, B color signals. Wideband progressive scan signal Y, I, Q is sampled at 8x chroma subcarrier rate (8xfsc) and separated vertical-temporal (VT) low pass filter in filter unit 16 to generate filtered signals YF, IF and QF Is converted from analog to digital (binary) form by a separate analog-to-digital converter (ADC) in the ADC unit 14 before each is filtered by These signals are shown as waveform A in FIG. The separation filters are 3x3 linear Shibu conversion filters of the type shown in FIG. 10d to be discussed below. These filters are designed to allow vertical inter-scanning of the main signal (component 1 in Figure 1) after progressive scanning for interlaced transformations to prevent undesirable interlacing artifacts (such as frickers, jagged edges, and other aliasing related effects). Temporal resolution Slightly reduces the diagonal VT resolution.

The center panel expansion factor (CEF) is a function of the difference between the width of the image displayed by the optical screen receiver and the width of the image displayed by the standard receiver. The image width of an optical screen display having an aspect ratio of 5: 3 is 1.25 times larger than that of a standard display having an aspect ratio of 4: 3. This 1.25 factor is the first center panel expansion factor that must be adjusted to know the overscan area of the standard receiver and to know some deliberate overlap of the border area between the center and side panels.

These considerations lead to consideration of the CEF of 1.19.

The progressive scan signal from the filter network exhibits a bandwidth of 0 to 14.32 MHz and is converted into a 2: 1 interlaced signal by the progressive scan P for the interlaced (I) converters 17a, 17b and 17c, respectively. This will be described with reference to FIGS. 22 and 23. The bandwidths of the output signals IF ', QF' and YF 'from the converters 17a to 17c represent a bandwidth of 0 to 7.16 MHz since the horizontal scan rate for the interlaced signal is half the horizontal scan rate of the progressive scan signal. . In the conversion process, the progressive scan signal is subsampled to take half of the effective pixel sample so that a 2: 1 interlaced main signal is generated. In particular, each progressive scan signal is converted to a 2: 1 interlaced scan format by reading the retained pixels at 4xfsc rate (14.32 MHz) or by retaining even or odd lines in each field. All continuous digital processing of the interlaced scanned signal is 4xfsc. Network 17c also includes an error prediction network. One output YF 'of network 17c is an interlaced subsampled luminance version of the prefiltered progressive scan component. The other output (luminance) signal YT of the network 17c contains temporal information derived from the video field difference information and represents a temporal prediction or temporal interpolation error between the actual value and the predicted value of the luminance sample missing at the receiver. This prediction is based on the temporal average of the amplitudes of the before and after pixels that are valid at the receiver.

The signal YT, i.e., the luminance helper signal that assists in reconstructing the progressive scan signal at the receiver, reveals the error that the receiver is expected to generate with respect to the unfixed video signal and facilitates the deletion of such error at the receiver. In the fixed part of the image, Era is zero and complete reconstruction is performed at the receiver. We found that chromatic helper signals are not necessary in practical matters, and that luminance helper signals are not enough for the human eye to detect lack of chromatic vertical or temporal detail, which is sufficient to produce good results.

Figure 2a shows the algebra used to generate the helper signal YT.

Referring to FIG. 2A, it can be seen that pixels A, X, and B in the progressive scan signal occupy equally spaced positions in the image. Black pixels such as A and B are transmitted as the main signal and are valid at the receiver. White pixels such as X are not delivered, which is predicted by the temporal frame average (A + B) / 2. In other words, the prediction of the missing pixel is made by averaging the amplitudes of the before and after pixels A and B in the encoder.

The predicted value A + B) / 2 is extracted from the actual value X corresponding to the helper signal having an amplitude according to the formula of X- (A + B) / 2 so that the predicted error signal is generated. This equation defines temporal field difference information in addition to temporal frame average information. The helper signal is horizontally low pass filtered by the 750 KHz low pass filter and passed as the helper signal YT. Band limiting of the helper signal to 750 KHz is due to the need to prevent this signal from colliding with the next lower RF channel after being modulated onto the RF image carrier.

At the receiver, similar prediction of the missing pixel X is performed by using the average of the samples A and B, and the prediction error is added to the prediction value. That is, X is restored to the temporal mean (A + B) / 2 by adding the prediction error X- (A + B) / 2. The helper signal thus facilitates the conversion from the interlaced scan format to the progressive scan format.

The helper signals produced by the described temporal predictive computations can be compared with the ENTSC 2-comparable to the IEEE Report HDTV system on CE-33, No. 3, pp. 146-153, August 1987. In an article called Channel. It is a low energy signal compared to the prediction signal produced by some other operation as used to generate a line differential signal as described by Mr. Tshinberg. In the image area, the error energy is zero because the prediction is complete. The energy content of the helper signal indicates whether the video signal remains stationary or moves image information. The low energy helper signal state is clarified by still and substantially still images (such as news broadcasts characterizing reporters for still backgrounds), while high energy states indicate image movement. The described operation has been found to produce at least opposing artifacts after image reconstruction at the receiver. The helper signal generated by the described operation retains its usefulness and is then band limited (filtered) to 750 KHz. The helper signal generated by the described operation advantageously represents zero energy when there is still image information. In conclusion, helper signals related to still images are not affected by filtering.

The temporal prediction system described is used in interlaced systems and advancing scans that are higher than the standard line rate. However, this system works best with an advancing scanning source with pixels A, X, and B occupying the same spatial location in the image. This temporal prediction will be incomplete at the still part of the image, even if the native optical screen image is from an interlaced signal source. In such a case, the helper signal will have more energy and will introduce a small artifact at the still part of the reconstructed image.

Experiments have shown that the use of interlaced signal sources yields acceptable results with artifacts that can only be recognized by closer examination. However, progressive scan signal sources introduce fewer artifacts and produce good results.

Referring back to FIG. 1A, the optical screen signals IF ', QF', YF 'interlaced from the transducers 17a-17c are each of the horizontal low pass filters 19a, 19b, 19c, i.e., 0-5 MHz. It is filtered by the horizontal low pass filter which produces a signal YF with a bandwidth, a signal QF with a bandwidth of 0 to 600 KHz and a signal IF with a bandwidth of 0 to 600 KHz.

This signal is subject to a format encoding process that encodes each of these signals in a 4: 3 format by means of a side-center signal separator and a format encoder coupled to processor unit 18.

In short, the central portion of each optical screen line is mapped and timed to the displayed portion of active line time according to an aspect ratio of 4: 3. Time extension reduces the bandwidth so that the inherent optical screen interlaced frequency is compatible with the standard NTSC bandwidth. The side panel is divided into horizontal frequency bands so that the color highs components of I and Q represent a bandwidth of 83KHz to 600KHz (as shown by the IH signal in FIG. 7), and the Y luminance highs component is 700KHz to 5.0MHz. The bandwidth of (shown as YH signal in FIG. 6) is shown. Side panel lows, e.g., YO, IO and QO signals developed as shown in Figures 6 and 7, have a DC component, and horizontal image light on the left and right along each line It is mapped to the screen area and time compressed. Side panels are treated separately. The contents of this format encoding process are as follows.

In considering the following encoding contents, it may also be helpful to consider FIG. 1e depicting the processing of the encoding components (1, 2, 3, 4) in the context of the displayed center and side panel information. Filtered interlaced signals (IF, QF, YF) are processor 18 and side-center panel generating three groups of output signals (YE, IE, QE and YO, IO, QO and YH, IH, QH). Processed by the signal separator. The first second group of signals YE, IE, QE and YO, IO, Q are processed to develop a signal containing the side panel luminance lows compressed into the horizontal light scan region and a central panel component of full bandwidth. .

The third group of signals YH, IH, QH is processed to develop a signal containing side panel highs. When these signals are combined, an NTSC compatible optical screen signal is generated with a 4: 3 display aspect ratio. The content of the circuit comprising the unit 18 is known and will be described in conjunction with FIGS. 6, 7, 8.

Signals YE, IE, QE contain complete center panel information and represent the same format as indicated by signal YE in FIG. In short, the signal YE is derived from the signal YF as follows. The optical screen signal YF holds side and center panel information and contains pixels 1 to 754 which occur during the active line intervals of such optical screen signals. This wide band center panel information (pixels of 75 to 680) is extracted as the center panel luminance signal YC through time demultiplex processing. Signal YC is time extended by a center panel expansion factor of 1.19 (e.g., 5.0 MHz to 4.2 MHz) that produces an NTSC compatible center panel signal YE.

Signal YE represents the NTSC compatible bandwidth (0 to 4.2 MHz) due to time extension by factor 1.19. The signal YE pictorializes the image display interval TD (FIG. 2) between the light scan areas T0. The signals IE and QE are each developed from the signals IF and QF, and are similarly processed by the method of the signal YE.

The signals (YO, IO, QO) provide low frequency side panel information (rose) inserted into the left and right horizontal overscan area. Signals YO, IO, QO represent the format as indicated by signal YO in FIG. In short, the signal YO is derived from the following signal YF. The optical screen signal YF contains left panel information related to pixels 1 to 84 and right panel information related to pixels 671 to 754.

As will be mentioned, the signal YF is obtained from the luminance low-low signal having a bandwidth of 0 to 7009 Hz from the signal from which the left and right side panel low-low signals are extracted through time demultiplexing (YL 'signal in FIG. 3). Low pass filter to generate. The luminance low signal YL 'is time-compressed to generate the side panel low signal YO along with the low frequency information compressed in the overscan area associated with pixels 1-14, 741-754. The compressed side lows signal exhibits an increased BW proportional to the amount of time compression. The signals IO and QO are each developed from the signals IF and QF and processed similarly to the signals of YO.

Signals (YE, IE, QE and YO, IO, QO) are side-centered signal combiners 28, such as time multiplexers, that produce signals having a 4: 3 aspect ratio and NTSC compatible bandwidth (YN, IN, QN). Combined by This signal consists of the signal YN shown in FIG. The combiner 28 also has a suitable time delay to equalize the transmission time of the combined signal. Such equalization signal delay is retained elsewhere in the system necessary to equalize the signal transmission time.

Modulator 30, band pass filter 32, H-V-T band stop filter 34 and combiner 36 constitute an improved NTSC signal encoder 31. Chromaticity signals IN and QN are orthogonally modulated on subcarrier SC at NTSC chromaticity subcarrier frequency, which is generally 3.58 MHz, by modulator 30 generating modulated signal CN.

The modulator 30 is designed in a conventional manner and will be described in connection with FIG.

The modulated signal CN is bandpass filtered in a vertical (V) and time (T) area by a two-dimensional (VT) filter 32 and applied as a signal CP to the chromaticity signal input of the combiner 36. Remove crosstalk from the previously interlaced chroma signal.

The luminance signal YN is horizontal (H), vertical (V) and time (T) area by the three-dimensional (HVT) band stop filter 34 before being applied as the signal YP to the luminance input of the combiner 36. In the band stop filter.

The filtering luminance signal YN and the chromatic color difference signals IN and QN allow the luminance-chromatic crosstalk to be significantly reduced later after NTSC encoding. Multi-area space-time filters, such as H-V-T filter 34 and V-T filter 32 in FIG. 1, comprise the structure as illustrated in FIG.

In FIG. 1A, the H-V-T band stop filter 34 shows the configuration in FIG. 10B, and removes the upper moving diagonal frequency component from the luminance signal YN. This frequency component is similar in appearance to the chroma subcarrier component and eliminates holes into which the modulated chromaticity is inserted in the frequency spectrum. The removal of the upward shifting diagonal frequency component from the luminance signal YN does not visually lower the determined displayed picture since the human eye is not really sensitive to this frequency component. Filter 34 exhibits a cutoff frequency of approximately 1.5 MHz in order not to corrupt the luminance vertical content information.

V-T band pass filter 32 allows modulated chromaticity side panel information to be inserted into holes made in the luminance spectrum by filter 34 to reduce chromatic bandwidth.

The filter 32 reduces the vertical and temporal resolution of the chromaticity information, and such stationary and moving edges are slightly blurred, but this effect is due to or little affected by the insensitivity of human vision to such effects.

The output center / side rose signal (C / SL) from the combiner 36 holds NTSC compatible information to be displayed and is located in the left and right horizontal overscan areas not visible by the observer of the NTSC receiver display, It is derived from the center panel of the optical screen signal as well as the compressed side panel lows (luminance and chromaticity) derived from the side panel of the screen signal. The side panel rose compressed in the overscan area represents one component of the side panel information for the optical screen display. Side panel heights, ie other components, are developed by the processor 18 as will be described later. The side panel high signal (YH; luminance high, IH: I high and QH; Q high) is illustrated in FIG. 6, 7, 8 illustrate a device for developing such a signal as will be described. In FIG. 4, signals YH, IH, and QH hold left panel high frequency information associated with left panel pixels 1 through 84, and the right panel high frequency information associated with right panel pixels 671 through 754. Holds.

The central panel portion of the signal C / SL is processed by the previously mentioned adjusting intra frame processor 38 which generates a signal N applied to the input of the adder 40.

The intra framed signal N is basically the same as the signal C / SL because of the high visual interrelationship of the intra frame image information of the signal C / SL. The image shift processor 38 averages the signals C / SL above approximately 1.5 MHz and helps to reduce or eliminate vertical-time artifacts between the main and peripheral signals. The processor 38 operates, and a high pass frequency range of 1.5 MHz is selected such that full intra frame averaging is achieved for information at 2 MHz, and luminance vertical content information is not compromised by the processing of intra frame averaging. The horizontal workpiece is removed by a 200 KHz supervisory band between the filter associated with the intra frame processor unit in the decoder of FIG. 13 and the filter associated with the intra frame averager 38 in the encoder 31. FIG. 11B shows the contents of the highs intra frame processor 38, and FIG. 11B and FIG. 13 will be described later.

The signals IH, QH, YH are located in NTSC format by an NTSC encoder 60 similar to the encoder. In particular, encoder 60 is shown in FIG. 9 as well as an apparatus for orthogonally modulating the side panel chroma height information on the side panel luminance height information at 3.58 MHz to generate a signal NTSCH in NTSC format. It has a device of the old form.

This signal is illustrated in FIG.

The use of multi-area bandpass filtering in NTSC encoders 31 and 60 favors chromatic and luminance components such that when the receiver has conservative multi-area filtering to separate the luminance and chromaticity information, it separates the workpiece freely vertically at the receiver. Allow it. The use of conservative filters for luminance / chromatic encoding and decoding is called collaborative processing, and collaborates for enhanced chromatic / luminance separation published in SMPTE Journal, Aug. 1986, Vol. 95, No. 8, pages 782-789. Mr. H. in an article under the disposition. It is explained in detail by Mr. Stroll. Standard receivers using existing notches and line-comb filters will ease the use of such multi-area pre-filtering in encoders by exhibiting reduced chromatic / luminance artifacts.

The signal NTSCH is time extended by unit 62 generating an extended side height signal EHS having an active horizontal line spacing of 50 less than approximately 52 standard NTSC active line spacings. In particular, as shown in FIG. 5, this extension extends to the pixel position 15 to 15 of the signal EHS, such as the left side elevation of the signal NTSCH extended to occupy approximately one-half the line time of the signal EHS. This is accomplished by a mapping process that maps the left side panel pixels 1-84 of the signal NTSCH to 377.

The right side panel portion 671-754 panels of the signal NTSCH are similarly processed. The time extension process reduces the horizontal bandwidth of the information containing the signal (ESH: compared to the NTSCH signal) by a factor of 363/84. The mapping process by means of the time extension achieved can be realized by the apparatus in the form shown, and will be described in connection with the twelfth to 12d degrees. The signal ESH is suitably intra-framed by the network 64 which generates the signal X as illustrated in FIG. 5 of the form shown in FIG. 11A. The intra frame averaged signal X is basically the same as the signal ESH because of the high visual correlation of the intra frame image information of the signal ESH. Signal X is applied to the signal input of quadrature modulator 80.

Signal YF 'is also filtered by horizontal band pass filter 70 with a pass band of -6.0 MHz. The output signal from the filter 70 and the horizontal luminance rise is applied to an amplitude modulator 72 whose amplitude is modulated with a 5 MHz carrier signal Fc. The modulator 72 has an output low pass filter with a cutoff frequency of approximately 1.0 MHz, which results in a signal with a 0 to 1.0 MHz passband at the output of the modulator 72. The upper (aliased) side band (5.0-6.0 MHz) generated by the modulation process is removed by a 1.0 MHz low pass filter. Effectively, in the range of 5.0 MHz to 6.0 MHz, the horizontal luminance rise frequency is shifted to the range of 0 to 1.0 MHz as a result of subsequent low pass filtering and amplitude modulation processing. The carrier amplitude must be large enough so that the natural signal amplitude remains after filtering by a low pass filter of 1.0 MHz. In other words, a frequency shift is generated without affecting the amplitude.

The horizontal luminance rise signal frequency shifted from the unit 72 is encoded (time compressed) by the format encoder 74.

In other words, the encoder 74 checks for a frequency shifted horizontal luminance rise such that this signal exhibits 50 active line spacing less than the standard NTSC active line spacing of 52.6 by using the technique described in connection with FIGS. Code it.

When the input signal to encoder 74 is time compressed by encoder 74, the bandwidth is increased from approximately 1.0 MHz to 1.1 MHz at the output of encoder 74. The signal from encoder 74 is properly intra framed by a device 76 similar to that illustrated in FIG. 11A before being applied to unit 80 as signal Z.

The intra frame averaged signal is basically the same as the signal from encoder 74 because of the high visual correlation of intra frame image information of the signal from encoder 74. The modulated signal X, the composite signal carrying the luminance and chromaticity information, and the modulated signal Z actually exhibit a bandwidth of approximately 0 to 1.0 MHz.

As will be explained in FIG. 24, the unit 80 performs nonlinear gamma function amplitude compression on the large amplitude congestion of the two peripheral signals (X, Z) before these signals orthogonally modulate the alternate subcarrier signal (ASC). do. A gamma of 0.7 is used, where the absolute value of each sample is raised to 0.7 power and multiplied by the sine of the intrinsic sample value. Gamma compression reduces the visibility that hinders large amplitude congestion of the modulated signal on an existing receiver, and because the inverse of the gamma function used at the encoder can be easily implemented and predicted at the receiver decoder, It allows you to predict recovery.

The amplitude compressed signal is then orthogonally modulated on a phase controlled alternative subcarrier (ASC) of 3.1075 MHz, an odd multiple of 1/2 of the horizontal frequency (395 x H / 2). The phase of an alternate subcarrier, unlike the phase of a chroma subcarrier, allows 180 ° to be replaced from one field to another. The field-substituted phase of the alternate subcarrier allows the peripheral modulation information of the signal (X, Z) to overlap with the chromaticity information and facilitates the separation of the surrounding information using a relatively uncomplicated field storage at the receiver. Peripheral information components A1, -A1, A3, -A3 with conservative phases are generated. In the quadrature modulated signal M, the signal N is added to the adder 40. The resulting signal NTSCF is an NTSC compatible signal of 4.2 MHz.

The nonlinear gamma function used in the encoder for large amplitude compression is a component of nonlinear companding (compression-expansion) that also holds a conservative gamma function at the decoder of the optical screen receiver for amplitude expansion as will be described later. The described nonlinear companding system has been found to significantly reduce the impact of surrounding nonstandard information on standard information without visual impairment of the image due to noise effects. This companding system uses a nonlinear gamma function that simultaneously compresses non-standard optical screen high frequency information at the encoder with large amplitude congestion at the encoder along with the conservative nonlinear gamma function used to extend such high frequency information at the decoder. This result is a reduction in the amount of disturbance with existing standard video information generated by large amplitude ambient high frequency information in the described compatible optical screen system. Here, the non-standard ambient light screen information is divided into a high frequency portion and a low frequency portion dependent on companding. At the decoder, the nonlinear amplitude extension of the compressed high frequency information is not overly aware of noise because large amplitude high frequency information is typically associated with high contrast image edges and human vision is not sensitive to noise at such edges. The described companding process also reduces the cross-modulation product between the alternate and chroma subcarriers with the associated reduction in the visual bit product.

The luminance content signal YT represents a bandwidth of 7.16 MHz and is encoded in a 4: 3 format by a format encoder 78 (in the method shown in FIG. 6), and generates a filter 79 for generating the signal YTN. Low pass filter horizontally at 750KHz.

The side portion has a cutoff frequency of 125 KHz but is low pass filtered to 125 KHz before being time compressed by an input low pass filter of the format encoder 78 matching the input filter 610 of the apparatus shown in FIG. The high side part is discarded.

So signal YTN correlates with main signal C / SL.

The signals YTN and NTSCF are converted to digital (binary) by analog by the DAC units 53 and 54, before which these signals are supplied to the RF quadrature modulator 57 to modulate the TV RF carrier signal. The RF modulated signal is applied to the transmitter to communicate via antenna 56.

Alternating subcarriers associated with modulator 80 are horizontally synchronized and select frequencies to maintain a moderate separation of center information and side ends (e.g., 20-30db) and to have a subtle effect on the image displayed by a standard NTSC receiver. . The ASC frequency will interlace the frequency at an odd multiple of the 1/2 horizontal line ratio so as not to cause interference that compromises the well displayed picture quality.

Orthogonal modulation as provided by unit 80 allows two narrowband signals to be transmitted simultaneously. The time to extend the high modulated signal results in a bandwidth reduction that matches the narrowband condition of quadrature modulation. The further the bandwidth is reduced, the less interference will result between the carrier and the modulated signal. In addition, the high energy DC component of the side panel information is typically reduced to an overscan area different from that used with the modulation signal. Thus, the energy of the modulated signal and the potential interference of the modulated signal are greatly reduced.

Encoded by the antenna 56 to cope with optical widescreen signal broadcast, the NTSC is received by the NTSC receiver and the optical screen receiver as illustrated by FIG.

In FIG. 13, a broadcast optical screen EDTV interlaced television signal is received by an antenna 1310 and supplied to an antenna input of an NTSC receiver 1312. Receiver 1312 is a portion of the modulated alternating subcarrier signal that does not interfere with standard receiver operation (e.g., high) and light that is forced (e.g., low) into the horizontal overscan area outside of the observer's field of view. The dual-side optical screen signal is processed in a normal form to generate screen display with screen side panel information and a 4: 3 aspect ratio.

A compatible optical screen EDTV signal received by the antenna 1310 is supplied to an optical screen sequential scanning receiver 1320 capable of displaying a video image with an optical aspect ratio of, for example, 5: 3. The received optical screen signal includes an input unit 1322 comprising a radio frequency (RF) tuner and an amplifier circuit, a synchronous video demodulator (orthogonal demodulator) for generating a baseband video signal, and a baseband video signal (NTSCF) in binary form. Is processed with an analog-to-digital (ADC) converter to generate The ADC circuit operates at four times the chroma subcarrier frequency (4xfsc) sampling rate.

The signal NTSCF is supplied to an intra frame processor 1324 which processes the image line 262H dropped within a frame of 1.7 MHz or more to reproduce a substantially free quadrature modulated auxiliary signal M and a main signal N with V-T crosstalk. The 200 KHz horizontal crosstalk guard band is provided between the low operating frequency of the unit 1324 and 1.7 MHz of the low operating frequency of the unit 38 in the encoder of FIG. 1A. The reproduced signal N includes information that matches the image information of the main signal C / SL due to the high visible intra frame image correction of the basic main signal C / SL such as the intra frame processed by the encoder of FIG.

Signal M is coupled to a quadrature demodulator and amplitude makeup unit 1326 to demodulate auxiliary signals X and Z in response to an alternating subcarrier ASC with a field alternating phase similar to the signal ASC described in connection with FIG. The demodulated signals X and Z are the intra frames processed by the encoder of FIG. 1a, and the same information as the output signal and the image information signal ESH from unit 74 in FIG. 1a is due to the high visible intra frame image correction of these signals. Include. Unit 1326 is a 1.5 MHz low pass filter, a non-linear compression implemented by unit 80 in Figure 1a, for example, with a reverse gamma function (gamma = 1 / 0.7 = 1.429) to remove undesirable high frequency demodulation. And an amplitude expander for extending the demodulated signal using the inverse of the function.

Unit 1328 time compresses the color encoded side panels such that they occupy their base time slots, thereby regenerating the signal NTSCH.

The unit 1328 time compresses the signal NTSCH to the same time as unit 62 of the first extended time signal NTSCH.

A high luminance (Y) decoder 1330 is introduced into the optical screen format with the time of extending the signal in the same time as the time compression of the corresponding component in the encoder of FIG. 1A using the mapping technique as shown in FIG. Decode the high luminance horizontal signal Z.

Modulator 1332 amplitude modulates a signal from decoder 1330 with a 5.0 MHz carrier fc. The amplitude modulated signal is high pass filtered by the filter with a 0.5 MHz cutoff frequency to remove less sidebands. For the output signal from the filter 1334, the center panel frequency of 5.0 to 6.0 MHz is reproduced, and the side panel frequency of 5.0 to 6.0 MHz is reproduced. The signal from filter 1334 is supplied to adder 1336.

Signal NTSCH from compressor 1328 is supplied to unit 1340 to separate it from high chromas to generate signals YH, IH, QH. This is completed by the configuration of FIG.

The signal N from the unit 1324 is separated into the constituent luminance and chromatic components YN, IN, QN by a luminance-chromat separator 1342 similar to the separator 1340 and using a device of the type shown in FIG.

The signals YH, IH, QH and YN, IN, QN are provided as inputs to the Y-I-Q format decoder 1344 to decode the luminance and chromaticity components into the optical screen format. The low side panels are time extended, the center panel is time compressed, the high side panels are added to the low side panels and the side panels are added to the center panel in a 10 pixel overlap area using the principle of FIG.

The decoder 1344 is shown in FIG. 19.

Signal YF 'is coupled to adder 1336 and sums up with signal from filter 1334. The high frequency horizontal luminance detail information reproduced and expanded by this process is added to the decoded luminance signal YF '. The output signal from the adder 1336 is carried to intersect the sequential scan converter 1350 via an adaptive frame repetition network 1335 responsive to the representative helper signal YT. Network 1335 passes the output signal from adder 1336 to converter 1350 as indicated in the condition of signal YT.

However, the image information network 1335 performs a frame repetition operation on the video signal 1336 for the optical screen format frequency from 1.78 MHz to 3.7 MHz, before which the signal is supplied to the converter 1350. The frame repetition process is illustrated by FIG. 26 as described above. Items of network 1335 are shown in FIG. 27.

Signals YF ', IF', QF 'are interlaced and converted into sequential scan formats by converters 1350, 1352, 1354. The luminance sequential scan converter 1350 also responds to the helper luminance signal YT from the format decoder 1360, which decodes the encoded helper signal YTN. Decoder 1360 decodes signal YIN into the optical screen format and exhibits a configuration similar to that of FIG.

I and Q converters 1352 and 1354 generate absent sequential scan line information, which interlaces one frame into a sequential scan signal by means of an average line. This will be completed with the device of the type shown in FIG.

The luminance sequential scan converter unit 1350 is similar to that shown in FIG. 20 except that the signal YT is added as shown by the arrangement in FIG. The helper signal sample YT in this unit is added as a temporary average to assist in reconstructing sequential scanning pixel samples. The entire temporary item is reproduced in the band of the horizontal frequency contained in the encoded line difference signal (750 KHz after encoding). The band of the horizontal frequency signal YT is zero, so the absent sample is reconstructed by a temporary average.

The optical screen sequential scanning signals YF, IF, QF are converted into analogue by the digital-to-analog converter 1362 before being supplied to the video signal processor and the matrix amplifier unit 1164. The video signal processor component of unit 1164 includes signal amplification DC level shifting, peaking, brightness control, contrast control, and other conventional video signal processing circuits. The matrix amplifier 1164 combines the luminance signal YF with the color difference signals IF and QF to generate color images of representative video signals R, G, and B. These color signals are amplified by the display drive amplifiers in the unit 1364 at a level suitable for directly driving the optical screen color image display element 1370, for example the optical screen kinescope.

FIG. 6 illustrates the apparatus included in the processor 18 of FIG. 1a to generate signals YE, YO, YH from the wide band optical screen signal YF. The signal YF is horizontally lowpass filtered by the input filter 610 at a cutoff frequency of 700 KHz to generate the low frequency luminance signal YL, which is supplied to one input of the subtractor 612.

The signal YF is delayed by the unit 614 to compensate for the signal processing delay of the filter 610 and then supplied to the time demultiplexing device 616 and to the other input of the subtractor 612. The sum delayed signal YF and the filtered signal YL generate a high frequency luminance signal YH at the output of the subtractor 612.

Delayed signal YF and signals YH and YL are supplied to branch inputs of demultiplexing device 616 including demultiplexing (DEMUX) units 618, 620, 621 for processing signals YF, YH and YL, respectively. The items of demultiplexing apparatus 616 will be described in conjunction with FIG.

The demultiplexing units 618, 620, 621 include the full bandwidth center panel signal YC, the high side panel signal YH, as illustrated in FIGS. 3 and 4; The low side panel signal YL 'is driven.

Signal YC is time extended by time expander 622 to generate signal YE. Signal YC is time extended with a center expansion factor sufficient to leave the room for the left / right horizontal overscan area. The central expansion factor 1019 is the ratio of the intended width of the signal YE (pixels 15 to 740) to the width of the signal YC (pixels 75 to 680) as shown in FIG. 3.

The signal YL 'is compressed at the side compression ratio by the time compressor 628 to generate the signal YO. The side compression ratio 610 is the width of the corresponding portion of the signal YL '(e.g., left pixels 1 to 84) at the intended width of the signal YO (e.g. left pixels 1 to 14) as shown in FIG. Ratio. Time expanders 622, 624, 626 and time compressor 628 may be configurable in the form shown in FIG. 12 as described.

Signals IE, IH, IO and QE, QH, QO are generated from signals IF and QF, respectively, in a similar way that signals YE, YH and YO are generated by the apparatus of FIG. In this regard, Figure 7 illustrates an apparatus for generating signals IE, IH and IO from signal IF. Signals QE, QH and QO are generated from signal QF in a similar manner.

In FIG. 7, after being delayed by the unit 714, the wideband optical screen signal IF is coupled to the demultiplexing device 716 and the low frequency signal from the low pass filter in the subtractor 712 to form the high frequency signal IH. Bound to IL.

Delayed signals IF and signals IH and IL are demultiplexed by demultiplexers 718, 720, 721, respectively, associated with demultiplexing device 716 to generate signals IF, IH, and IL '.

Signal IC is time extended by expander 722 to generate signal IE and signal IL 'is time compressed by compressor 728 to generate signal IO. The signal IC is expanded with a central expansion rate similar to that used for signal YC as recorded, and the signal IL 'is compressed with a side compression ratio similar to that used for signal YL'.

FIG. 8 illustrates a demultiplexing device 816 such as may be used for the device 616 of FIG. 6 and the device 716 of FIG. The apparatus of FIG. 8 is illustrated with the contents of the demultiplexer 616 of FIG. The input signal YF includes 754 pixels that define the image information. Pixels 1 through 84 define the left panel, pixels 671 through 754 the right panel, and pixels 75 through 680 define a center panel that lightly overlaps the left / right panels. Signals IF and QF show similar overlap. The panel overlap to be described has found the ease of combining the center and side panels at the receiver to almost eliminate the border.

The demultiplexing apparatus 816 includes first, second, and third demultiplexer (DEMUX) units 810, 812, 814, respectively, associated with left, center, and right panel information.

Each demultiplexer unit has an input A to which signals YH, YF and YL are supplied and an input B to which a blanking signal BLK is supplied. The blanking signal will be at logic zero level or ground. The unit 810 receives a first control signal from a coefficient comparator 817 in which the signal selection input SEL of the unit 810 indicates the presence of left panel pixel components 1 to 84 and right panel pixel components 671 to 754. While output signal YH including left / right highs is derived from input signal YH.

The second control signal from coefficient comparator 817 also generates a BLK signal at input B that is different from signal YH at input A to be coupled to the output of unit 810. Unit 814 and coefficient comparator 820 operate in a similar fashion to drive the low side panel signal YL 'from signal YL. Unit 812 combines the signal YF from the input of the unit to the output of the unit to generate the center panel signal YC only when the control signal from coefficient comparator 818 represents the actual of center panel pixels 75-680.

Coefficient comparators 817, 818, and 820 receive video signals YF by pulse output signals from counter 822, which respond to clock signals at quadruple luminance subcarrier frequencies (4xfsc) and in response to horizontal line sync signals H derived from video signals YF. Are synchronized. Each output pulse from counter 822 corresponds to a pixel position along a horizontal line. Counter 822 shows an initial offset of -100 coefficients where time pixel 1 is at the onset of the horizontal line display period and represents 100 pixels from the beginning of the negative incoming horizontal sync pulse to the end of the horizontal blanking period in time. Thus, counter 822 represents a coefficient of one in the on set of line display periods.

The principle used by the demultiplexing device 816 will be supplied to the multiplexing device to perform a signal summing operation such as that performed by the side-center panel summer 28 in FIG. 1A.

9 illustrates the breakdown of modulator 30 at encoders 31 and 60 of FIG. In FIG. 9, signals IN and QN represent four times the chroma subcarrier ratio 4cfsc and are supplied to signal inputs of latches 910 and 912, respectively. The latches 910 and 912 also receive a 4xfsc clock signal, a 2xfsc switching signal applied to the inverted switching signal input of the latch 910 and a non-inverting switching signal input of the latch 912 for transmission to the signals IN and QN. The signal outputs of latches 910 and 912 are combined into a single output line where signals I and Q appear alternately and are applied to the signal inputs of non-inverting latch 914 and inverting latch 916. These latches are clocked at 4fsc rates and receive switching signals at the chroma subcarrier frequencies fsc, inverted and non-inverted inputs. Non-inverting latch 914 generates an output alternating sequence of positive signals I and Q, and inverting latch 916 generates an output alternating sequence of negative polarities I and Q, eg, -1, -Q. The outputs of the latches 914 and 916 are combined in a single output line representing alternating sequence signals, such as 1, 0, -1, -Q, ..., etc., which constitute the signal CN. The signal is converted into a filtered signal of luminance signal YN to generate NTSC encoded signal C / SL such as Y + l, Y + Q, Y-1, YQ, Y + l, Y + Q. Filtered by filter 32 prior to coupling in 36).

FIG. 10 illustrates a vertical-temporal (V-T) filter that may represent a VT band pass, V-T band stop, or V-T low pass configuration by adjusting the weighting coefficients a1 to a9.

The table in FIG. 10A illustrates the weighting coefficients associated with the V-T band pass and band stop filter configurations used in the described system. HVT band stop filter, such as filter 34 of FIG. 1A, HVT band pass filter, such as that included in decoder system of FIG. 13, VT band stop filter 1021 and horizontal low pass filter (shown in FIG. 1020, a combination of a VT band pass filter 1031 and a horizontal band pass filter 1030, respectively, as shown in FIG. 10C.

In the H-V-T band stop filter of FIG. 10B, the horizontal low pass filter 1020 exhibits a given cutoff frequency and provides filtered low frequency signal components. This signal is summed by subtracting in summer 1023 with a delay of the input signal from delay unit 1022 to generate a high frequency signal component. The low frequency component is subject to one frame delay by the network 1024 before being applied to the adder 1025 to provide an H-V-T band stop filtered output signal.

V-T filter 1021 represents the V-T band stop filter coefficient shown in FIG. 10A. An HVT band pass filter as included in the decoder in FIG. 13 is horizontal containing a cutoff frequency given in dependence with a VT band pass filter 1031 having a VT band pass filter coefficient as indicated by the table of FIG. 10a. Shown in FIG. 10C as having a band pass filter 1030.

The filter of FIG. 10 provides a continuous signal delay at each tap t1 to t9 and includes a number of dependent memory units M 1010a to 1010h to provide the total filter delay. The signal carried by the tap is fed to multiplexers 1012a to 1012i of one input, respectively. The other inputs to each multiplexer each receive the weights a1 to a9 described as the features of the filtering process are performed. The nature of the filtering process also indicates the delay set by units 1010a to 1010h. The horizontal size filter uses pixel memory memory elements such that the total filter delay is less than the time period of one horizontal image line 1H. The vertical size filter uses the line memory memory component and the temporary size filter uses the frame memory memory component. Thus, the H-V-T 3-D filter has a combination of pixel 1H, line 1H and frame 1H storage components, and the V-T filter has only the latter two forms of memory components. The tap signal weighted from components 1012a-1012i is combined at adder 1015 to generate a filtered output signal.

Such a filter is a non-repeating, finite impulse response (FIR) filter. The nature of the delay provided by the memory component depends on the type of signal to be filtered and the amount of crosstalk between the luminance, chromaticity and high side channel signal in this embodiment. The clarity of the filter cutoff characteristic is improved by increasing the number of cascaded memory elements.

FIG. 10D illustrates generating the output signal and associated multipliers 1042a to 1042e according to the respective weight coefficients a1 to a5 designated for receiving signals from the cascaded memory (delay) units 1040a to 1040d, the signal taps t1 to t5. 1 shows a separate filter of the network 16 of FIG. 1a including a signal combiner 1045 which combines the output signals weighted from multipliers a1 to a5.

FIG. 11A illustrates a mobile adaptive intra frame processor suitable for use as units 64 and 76 of FIG. 1A. The input composite video signal is applied to a delay network comprising 262H delay elements 1102 and 1104 according to the associated input, output and intermediate tap terminals a, c and b. Signals from terminals a, b, and c are applied to each signal input of field repeat multiplexer (MUX) 1106, and signals from terminals a and c are applied to each signal input of field average MUX 1108. The MUX 1108 is switched at a field rate in response to a 30 Hz switching signal SW that is vertically synchronized in response to a vertical period sync pulse associated with the input composite video signal. MUX 1106 switches the input (labeled 0, 1, 2) to an output in response to switching control signals SW1 and SW2. Signals SW1 and SW2 are passed from the field identifier logic control circuit of a conventional design such that input 1 of MUX 1106 is coupled to the output in fields 1 and 4 and input 0 is coupled to the output in field 2. , Input 2 is coupled to the output in field 3. The output signal from MUX 1106 and the intermediate tap signal from terminal b are combined by network 1110 after being counted by a positive and negative single weight coefficient, respectively. The output signal from the MUX 1108 and the middle tap signal are combined by the network 1112 after being weighted by a positive and negative average coefficient of 1/2. The weight coefficients may be provided by a suitable matrix network in the combination network or by a signal multiplier of the input signal path of the combination network.

Output signals from combiners 1110 and 1112 are applied to signal inputs of MUX 1115 that receive the movement adaptive process control signal MAP in the switching control signal. The combiner 1120 is filtered by the band pass filter 1116 with a 1.5 MHz to 3.1 MHz pass band and then combines the output signal from the MUX 1115 and filtered by the high pass filter 1118 with a cutoff frequency of 3.1 MHz. Then combine the outputs of combiner 1112. The output signal from combiner 1120 is combined by network 1125 with an intermediate tap signal from terminal b after being dependent on delay circuit 1127 equalizing the transition time of the combined signal by network 1125. .

The apparatus of FIG. 11A shows either the field repetition operation in the case of a still image or the field average operation in the case of a moving image in response to a control signal MAP applied to the MUX 1115.

The control signal MAP, preferably a binary signal, is drawn out of the helper signal by a detector 1130 that includes a signal state indication threshold comparison circuit for sensing when the magnitude of the signal YT is displayed as image movement. MUX 1115 is enabled, so that field repetition operation is performed when signal MAP represents a still picture.

MUX 1115 is disabled, so that field intra frame averaging is performed when signal MAP indicates a predetermined amount of image movement.

FIG. 11B illustrates a mobile adaptive intra frame processor suitable for use as the unit 38 of FIG. 1A.

The device of FIG. 11B is similar to the device of FIG. 11A except that the 1.5 MHz horizontal high pass filter 1140, the electron gate 1144, and the combiner 1146 are added to FIG. 11B.

The output signal from combiner 1125 is filtered by 1.5 MHz horizontal high pass filter 1140 before being applied to electron transfer gate 1144. Gate 1144 responds to the switching control signal to pass a high frequency signal from the output of filter 1140 only during the center portion (part 1) of the main signal. At this time, the gate 1144 opens (conducts).

Gate 1144 is closed (non-conductive) during the time-compressed side panel portion of the main signal, for example during the described constant pulse period of the control signal. The output signal from gate 1144 is summed in combiner 1146 with the composite video signal delayed from center tap terminal b. The gate control signal is vertically synchronized in response to a vertical interval sync pulse coupled with the input composite video signal. The gate control signal is also horizontally synchronized.

Horizontal synchronization is achieved by means including a pixel counter responsive to the horizontal line sync pulse component of the input composite video signal and determining the timing of the positive pulse component of the bait control signal following each horizontal line sync pulse. The predetermined time interval is easily set between the horizontal line sync pulse and the first image pixel.

FIG. 12 shows a raster mapping apparatus that can be used for the time stretchers and compressors of FIGS. 6 and 7. In this regard, reference is made to the waveform of FIG. 12A showing the mapping process.

FIG. 12A shows the input signal waveform S along the center between pixels 84 and 670 which are mapped into pixel positions 1 to 754 of the output waveform W by time decompression processing. Endpoint pixels 1 and 670 of waveform S map directly to endpoint pixels 1 and 754 of waveform W. Intermediate pixels map directly on a 1: 1 basis with time extension, and in many cases do not map on an integer basis.

The latter case is explained when the pixel position 85.33 of the input waveform S corresponds to the integer pixel position 3 of the output waveform W, for example.

Accordingly, the pixel position 85.33 of the signal S includes the integer portion 85 and the fractional portion DX (.33), and the pixel position 3 of the waveform W includes the integer portion 3 and the fractional portion 0.

In Fig. 12, the pixel counter operating at the 4xfsc rate provides an output write address signal M indicating the pixel position 1 ... 754 on the output raster. The signal M is applied to a PROM (Programmable Read Only Memory) 1212 containing a lookup table containing values programmed by the nature of the raster mapping to be performed, for example by compression or decompression.

In response to signal M, PROM 1212 provides an output read address signal N representing an integer and an output signal DX representing a fraction greater than or equal to 0 but less than one. In the case of the 6-bit signal DX (2 = 64), the signal DX represents fractions 0, 1/64, 2/64, 3/64 ... 63/64.

PROM 1212 decompresses or compresses the video input signal S as a function of the stored value of signal N. Thus, the programmed value of the read address signal N and the programmed value of the fractional signal DX are provided in response to the integer value of the pixel position signal M. To extend the signal, for example, PROM 1212 is arranged to generate signal N at a rate slower than signal M.

Conversely, to compress the signal, PROM 1212 generates signal N at a rate greater than signal M.

The video input signal S is the pixel delay elements 1214a, 1214b cascaded to generate video signals S (N + 2), S (N + l) and S (N), which are mutually delayed versions of the video input signal; Delayed by 1214c. This signal is applied to the video signal input of each dual port memory 1216a-1216d as is known. Signal M is applied to each write address input of memories 1216a to 1216d, and signal N is applied to each read address input of memories 1216a to 1216d. Signal M determines where the incoming video signal information is written into the memory, and signal N determines the value is read from the memory. Such a memory writes into one address and reads from another address at the same time. The output signals S (N-1), S (N), S (N + l) and S (N + 2) from the memories 1216a-1216d are timed by the read / write operations of the memories 1216a-1216d. Represents a decompressed or time compressed format, which is a function of how PROM 1212 is programmed.

Signals S (N-1), S (N), S (N + l) and S (N + 2) from memories 1216a-1216d are four point linear interpolation including peaking filters 1220 and 1222. Is processed by the PROM 1225 and the two point linear interpolator 1230, which are described in detail in FIGS. 12B and 12C. Peaking filters 1220 and 1222 not only receive the peaking signal PX as shown, but also signal groups including signals S (N-1), S (N), S (N + l) and S (N + 2). Receive three signals from The value of the peaking signal PX varies from 0 to 1 as a function of the value of the signal DX as shown in FIG. 12D and is provided by the PROM 1225 in response to the signal DX. PROM 1225 includes a lookup table and is programmed to indicate a predetermined value of PX in response to a predetermined value of DX.

Peaking filters 1220 and 1222 provide peak and mutually delayed video signals S '(N) and S' (N + l) to a two point linear interpolator 1230 that receives signal DX. The interpolator 1230 provides a video output signal (compressed or decompressed) where the output signal W is defined by the compressor.

W = S '(N) + DX [S' (H + l) -S '(N)]

The four-point interpolator and peaking function described are advantageously a sin x / X interpolation function with high frequency detail resolution.

12B is a detailed view of peaking filters 1220 and 1222 and interpolator 1230.

In FIG. 12B, signals S (N-1), S (N) and S (N + l) are peaking filters whose weights are respectively weighted by peaking coefficients -l / 4, l / 2 and -l / 4. Applied to weight circuit 1240 in 1220. As shown in FIG. 12C, the weight circuit 1240 sets the signals S (N-1), S (N) and S (N + 1) to the peaking coefficients -l / 4, l / 2 and -l /, respectively. And multipliers 1241a through 1241c for multiplying to four. The output signals from multipliers 1241a through 1241c are summed in adder 1242 to generate peak signal P (N), which is signal S in adder 1244 to generate peak signal S '(N). The signal is multiplied by the signal PX in the multiplier 1243 to generate a peak signal that is summed with (N).

Peaking filter 1222 exhibits similar structure and operation.

In the two-point interpolator 1230, the signal S '(N) is subtracted from the signal S' (N + l) in the subtractor 1232 to generate a difference signal that is multiplied by the signal DX in the multiplier 1234. . The output signal from multiplier 1234 is summed with signal S '(N) in adder 1236 to generate output signal W.

FIG. 15 is a detailed view of the intra frame processor 1324 of FIG.

The input composite video signal for processor 1324 of FIG. 15 includes signal components Y1 + C1 and M1 + A1 in the first field. In the second continuous field, the input signal includes components' Y2 + C2 and M1-A1. Components Y1 + C1, M1 and Y2 + C2, M1 are components provided by the intra frame processor 38. Components + A1 and -A1 represent optional subcarrier signals in which intra-framed information from units 64 and 76 is modulated into components 2 and 3 during each successive field. Reference is made in particular to the first, 1a and 1d degrees.

With the MUX 1525 in position 1, the field difference component is achieved at the output of the combined value 1528. After being filtered by high pass filter 1530 and gated by unit 1532, it is modulated to generate a restored main signal Y1 + C1, M1 when combined with signal Y1 + Cl in combiner 1534. Component -A1, which deletes the auxiliary subcarrier component (+ A1), is generated. Component Y1 + C1 of the reconstructed main signal does not change below the 1.7 MHz cutoff frequency of high pass filter 1530, and component M1 represents intra framed center panel information of approximately 1.7 MHz or more. After inverting by the single gain amplifier 1535, the field difference cancellation (-A1) is the recovered and modulated auxiliary signal A1.

The recovered main signals Y1 + C1, M1 correspond to the signal N in FIG. 13 and are processed by the network 1342 as described. The recovered auxiliary signal A1 corresponds to the signal M in FIG. 13 and is demodulated by the network 1326.

FIG. 16 shows the operation of the network 1324 during the next consecutive video field as shown in FIG.

In this case, the signals Y2 + C2, M1-A1 are generated between the delay elements 1520 and 1522, and the MUX 1528 receives a position 2 for receiving the signals Y1 + Cl, M1 + A1. The recovered main signal Y2 + C2, M1 is generated at the output of the combiner 1534 and the opposite phase modulated auxiliary signal -A1 is recovered.

In FIG. 18, the HVT band pass filter 1810 having a passband of 3.58 ± 0.5 MHz and the configuration of FIG. 10c passes the signal NTSCH to the subtractor combiner 1814, and also through the transition time equivalent delay unit 1812. Receive signal NTSCH after passing. Separated luminance high signal YH occurs at the output of combiner 1814. The filtered NTSCH signal from filter 1810 is orthogonally demodulated by demodulator 1816 in response to chroma subcarrier signal SC to generate chromaticity high signals 1H and 8H.

In FIG. 19, signals YN, IN and QN are compressed by side-center panel signal separator (time demultiplexer) 1940 into side panel low signal YO, IO, QO and extended center panel signals YE, IE, QE. Are separated. Demultiplexer 1940 may use the principles of demultiplexer 816 of FIG. 8 as described above.

The signals YO, IO, and QO are time-extended by the time stretcher 1942 (corresponding to the side compression coefficient in the encoder of FIG. 1a), as indicated by the recovered side panel low signals YL, IL, and QL. Restore the circumferential relationship of the side panel low signal in the optical screen signal. Similarly, in order to form a room for the side panels, the center panel signals YE, IE and QE are time compressed by the time compressor 1944 (corresponding to the center elongation coefficient in the encoder of FIG. 1a) and restored. Restore the main spatial relationship of the center panel signal in the optical screen signal as indicated by the center panel signals YC, IC and QC. Compressor 1944 and expander 1942 are of the type shown in FIG. 12 as described above.

The side panel high signals YH, IH and QH restored to space are combined with the side panel low signals YL, IL and QL restored to space by combiner 1946 to generate reconstructed side panel signals YS, IS and QS. This signal is spliced into the center panel signals YC, IC and QC reconstructed by the splicer 1960 to form the fully reconstructed optical screen luminance signal YF and the optical screen chrominance signals IF 'and QF'. Splicing of the side and center panel signal components removes visible seams at the interface between the center and side panels, as can be seen in the collateral technique of splicer 1960 shown in FIG. Is achieved.

In FIG. 20, the interlaced signal IF '(or QF') is delayed 263H by element 2010 before being applied to the input of dual port memory 2020. This delayed signal is likely to be additionally delayed by 262H by element 2012 before being added to the input signal in adder 2014. The output signal from the adder 2014 is coupled to the two split network 2016 before being applied to the input of the dual port memory 2018. Memory 2020 and 2018 read data at 8xfsc rate and write data at 4xfsc rate. Outputs from memories 2018 and 2020 are applied to multiplexer (MUX) 2022 to generate an output sequential scan signal IF (QF).

A waveform diagram of a sequential scan output signal consisting of pixel samples C and X and an interlaced input signal (two lines according to the specified pixel samples C and X) are shown.

FIG. 21 shows a device suitable for use as a converter 1350 for the signal YF in FIG. Interlaced signal YF 'is delayed by elements 2110 and 2112 before being combined in adder 2114 as shown. The delayed signal from element 2110 is applied to dual port memory 2120. The output signal from adder 2114 is coupled to a two split network 2116, and its output is added to signal YT in adder 2118. Output from adder 2118 is applied to dual port memory 2122. Memory 2120 and 2122 provide an output signal to multiplexer 2124 that writes at 4xfsc rate, reads at 8xfsc rate, and generates sequential scan signal YF.

14 shows a side panel-central panel splicing device suitable for use as, for example, the splicer 1960 of FIG. In FIG. 14, the splicer, as well as the I signal splicer 1420 and the Q signal splicer 1430, similar to the structure and operation of the network 1410, as well as the side panel luminance signal component YS and the center panel luminance signal component It is shown with a network 1410 that generates the full bandwidth luminance signal YF 'from YC. The center panel and side panels overlap with many pixels, for example ten pixels. Thus, the center and side panel signals share a number of residual pixels through transmission processing and signal encoding before splicing.

In an optical screen receiver, the center and side panels are reconstructed from each signal, but many pixels at the side and center panel boundaries are wrong or distorted due to the time stretching, time compression, and filtering performed on the panel signals. The overlap area OL and the wrong pixel CP (slightly exaggerated for clarity) are represented by the waveforms associated with the signals YS and YC in FIG.

If the panel does not have overlapping areas, the wrong pixels are in contact with each other and the seam is visible. The ten pixel width of the overlap region is wide enough to compensate for three to five wrong boundary pixels.

The remaining pixels advantageously blend the side and center panels in the overlapping area. The multiplier 1411 multiplies the side panel signal YS by the weight function W in the overlap region as shown by the associated waveform before the signal YS is applied to the signal combiner 1415. Likewise, multiplier 1412 multiplies the center panel signal YC by the complementary weight function 1-W in the overlap region as shown by the associated waveform before signal YC is applied to combiner 1415. This weight function exhibits a linear ramp-like characteristic through the overlap region and includes a value between 0 and 1.

After weighting, the side and center panel pixels are summed by combiner 1415 such that each reconstructed pixel linearly combines the side and center panel pixels.

The weight function preferably approaches 1 near the innermost boundary of the overlapping region and approaches 0 at the outermost boundary. This has relatively little effect on the panel boundary where the wrong pixel is reconstructed. The linear ramped weight function described meets this requirement. However, the weight function is not linear, and nonlinear weight functions near the curved or circular ends, i.e. 1 and 0 weight points, may be used. Such weight function can be easily accomplished by filtering the sharp ramp weight function of the described type.

The weight functions W and 1-W are easily generated by a network including a lookup table responsive to an input signal indicative of pixel position, and a subtractor. The side-center pixel overlap position is known and the lookup table is programmed to provide an output value of 0 to 1 corresponding to the weight function W in response to the input signal. The input signal is generated in various ways by bars and values by the counter synchronized with each horizontal line sync pulse. The complementary weight function 1-W is generated by subtracting the weight function W from one.

FIG. 22 shows a device suitable for use as a sequential scan to interlace the transducer 17c for the signal YF in FIG. 1a, and also as shown in FIG. 2a, the vertical (V) and temporal (T) planes. A diagram of some sequential scan input signals YF according to samples A, B, C and X in the diagram. The sequential scan signal YF is prone to a 525H delay through elements 2210 and 2212 to generate samples X and A that are relatively delayed from sample B.

Samples B and A are summed in adder 2214 before being applied to two split network 2216. The output signal from network 2216 is subtractively combined within network 2218 with sample X to generate signal YF. The signal YT is applied to the input of the dual port memory 2222, and the signal YF from the output of the delay unit 2210 is applied to the input of the dual port memory 2223. Two memories 2222 and 2223 read at 4xfsc rate and write at 8xfsc rate to generate interlaced signals YF 'and YT at each output.

FIG. 23 shows an apparatus suitable for use as the transducers 17a and 17b of FIG. 1a. In FIG. 23, the sequential scan signal IF (or QF) is read at 4xfsc rate to generate interlaced scan output signal IF '(or QF') and delayed 525H before being applied to dual port memory 2312 which writes at 8xfsc rate. Applied to device 2310. Also shown are waveforms showing sequential scan input signals along the first and second lines associated with samples C and X, and interlaced scan output signals (first line along sample C stretched at an H / 2 rate). The dual port memory 2312 is stretched and outputs only the first line sample C of the input signal.

24 is a detailed view of the unit 80. Signals X and Z are applied to address inputs of nonlinear amplitude compressors 2410 and 2412, respectively. Compressors 2410 and 2412 are each programmable read-only memory (PROM) device that includes a look-up table that includes programmed values corresponding to a given nonlinear gamma compression function. This function is described as a simultaneous input to output response adjacent to unit 2412. Compressed signals X and Z from the data outputs of units 2410 and 2412 are applied to the signal inputs of signal multipliers 2414 and 2416, respectively. The reference inputs of multipliers 2414 and 2416 receive each optional subcarrier signal ASC in a mutually orthogonal phase relationship, which is sine and cosine type. Output signals from multipliers 2414 and 2416 are added into combiner 2420 to generate quadrature modulated signals M. In the decoder device of FIG. 13, the compressed signals X and Z are reconstructed through a conventional quadrature demodulation technique, and the complementary nonlinear amplitude extension of such a signal is compared with a lookup table programmed to a value complementary to the values of PROMs 2410 and 2412. This is done by the associated PROM.

FIG. 27 is a detailed view of the adaptive frame repetition unit 1335 of FIG. The input signal from block 1336 of FIG. 13 is applied to one signal input of delay unit 2710 and multiplexer (MUX) 2714. The other input of the MUX 2714 receives a version of the input signal delayed by the units 2710 and 2712. MUX 2715 receives an output signal from MUX 2714 at one input and an intermediate tap signal drawn at a point between delay networks 2710 and 2712 at the other input. The middle tap signal and the output signal from the MUX 2715 are applied to filters 2730 and 2732, respectively. Output signals from filters 2730 and 2732 are combined to generate an output signal coupled to network 1350 of FIG.

Filter 2730 exhibits a low pass response for frequencies of DC to 1.78 MHz and a band pass response for frequencies of 3.7 to 5.0 MHz. Filter 2732 exhibits a band pass response for frequencies from 1.78 MHz to 3.7 MHz and a high pass response for frequencies above 5.0 MHz. The frequency response in the optical screen format is represented by the frequency response of the filter in the encoder device of FIGS. 11A and 11B which processes the encoded signal for transmission in the standard aspect ratio format.

The switching of the MUX 2715 is controlled in response to the output signal from the logic AND gate 2720 applied to the input SEL of the MUX 2715. The AND gate 2720 responds to the output signal from the logical OR gate 2722 and the movement adaptive processing signal MAP fetched by the detector 2724 from the luminance helper signal YT.

OR gate 2722 responds to field 2 identifier signal F2 and field 3 identifier signal F3. The identifier signal F2 is also applied to the switching control input SEL of the MX 2714.

MUX 2715 couples one signal input to the output whenever AND gate 2720 represents one output logic level. This occurs in the absence of image movement when the signal MAP represents one logical level, either the identifier signal F2 or F3 represents one logical level, and fields 2 or 3 are provided. Appropriate apparatus for generating the field identifier signal can be easily generated as described in the circuit concept text by Eastman, pages 88-92, Gerald A, available from, for example, Oregon, Beaverton, Tektronix. Can be.

Claims (3)

An apparatus for receiving a television-type input signal that has undergone intraframe processing, the apparatus comprising: a decoder circuit coupled to the television input signal to provide a control signal (YT) representing a video shift of the television signal; (1360); Coupled to the input television signal and in response to the video movement content control signal, the television signal, which is substantially unchanged when a predetermined amount of video movement content is present, is passed through and is indicated by the movement content control signal. A frame repeating circuit (1337) for repeating the frame of the television signal when there is considerably less video moving content than the predetermined amount of video moving content; Circuitry (1364) coupled to the frame repeating circuit (1337) for processing the television signal in the form of an image representative signal; Television input signal receiving apparatus comprising a. The apparatus of claim 1, wherein the frame repetition circuit (1337) performs the frame repetition over a frequency range above a frequency range occupied by vertical detail information. 2. The apparatus of claim 1, wherein the television signal is in interlaced form and the apparatus includes a converter 1350 for converting the interlaced form of signal into a progressive scanning form. A television input signal receiving device (1337) for coupling the television signal to the scan converter.

Images